Copy, Paste, Compromise: How 'InstallFix' Malware Hijacks AI Coding Tool Downloads

It starts with a simple search: you're eager to try out Claude Code, Anthropic’s powerful AI coding assistant. You click the top Google result, copy the recommended install command, and paste it into your terminal-just like the documentation says. But in seconds, your credentials are stolen, and your development environment is compromised. Welcome to the world of 'InstallFix,' where convenience is a hacker’s best friend.

Fast Facts

• 'InstallFix' is a new malvertising campaign targeting AI coding tools like Claude Code.
• Attackers use Google-sponsored ads to lure victims to fake install pages.
• Malicious commands copied from these pages install the Amatera Stealer malware.
• Both experienced developers and newcomers are vulnerable due to common install practices.
• Threat actors abuse legitimate hosting services to blend in with normal web traffic.

How the Trap is Set

Researchers from Push Security have uncovered a fast-moving threat campaign that leverages the rising popularity of AI-powered coding tools. By cloning the official install pages of Anthropic’s Claude Code, attackers create near-perfect fakes that are promoted via Google-sponsored ads. These malicious links are especially dangerous because they bypass traditional email security filters, appearing as harmless search results to unsuspecting users.

Once on the fake site, users are urged to copy a one-line command and paste it directly into their terminal-a practice now standard for installing many developer tools. But this shortcut comes at a steep price: the command silently downloads and runs Amatera Stealer, a piece of malware designed to harvest credentials and potentially open the door to much broader enterprise breaches.

“There was a time when blindly pasting commands from a website was unthinkable,” notes Jacques Louw, Push Security’s co-founder. Today, both seasoned coders and AI-curious newcomers fall into the habit, making this attack vector devastatingly effective. The campaign is not just targeting beginners; it’s exploiting anyone who trusts a domain at face value, a trust that is increasingly misplaced as attackers use legitimate services like Cloudflare Pages and Squarespace to host their traps.

The Bigger Picture

The 'InstallFix' operation signals a shift in how social engineering and malvertising intersect. Instead of phishing emails, attackers are turning to search ads-an overlooked vector that reaches users right at the moment they’re looking to install new tools. With AI adoption surging, a wider, less security-savvy audience is now at risk, and the old wisdom of “never paste random commands” has never been more urgent.

Push Security warns that the threat is evolving rapidly: malicious domains appear and disappear in hours, making traditional blacklists ineffective. The only real defense is user vigilance-double-checking URLs, verifying sources, and resisting the urge to take installation shortcuts.

Conclusion

As AI coding tools become mainstream, cybercriminals are poised to exploit every gap between convenience and caution. The 'InstallFix' campaign is a stark reminder that security hygiene-especially for something as simple as copying an install command-can make the difference between innovation and intrusion. In the race to adopt the latest tech, don’t let good habits be the first casualty.

WIKICROOK

• Malvertising: Malvertising is the use of online ads to spread malware, often by tricking users into clicking harmful links-even on trusted websites.
• Social Engineering: Social engineering is the use of deception by hackers to trick people into revealing confidential information or providing unauthorized system access.
• Command: A command is an instruction sent to a device or software, often by a C2 server, directing it to perform specific actions, sometimes for malicious purposes.
• Credential Stealer: A credential stealer is malware designed to locate and steal passwords, digital keys, or authentication tokens from a victim’s computer or device.
• Indicators of Compromise (IoCs): Indicators of Compromise (IoCs) are clues like filenames, IPs, or code fragments that help detect if a computer system has been breached.